Personalized design of clothing pattern based on KE and IPSO-BP neural network

Abstract

In order to improve the precision of clothing development of fast fashion brands, consumers’ sense of experience, and brand loyalty, a design method of clothing pattern is proposed by combining Kansei engineering theory and improved particle swarm optimization (IPSO)–back propagation neural network (BPNN) model. First, based on the theory of Kansei engineering, the perceptual image experiment of clothing patterns was designed, and the mean value of perceptual image evaluation of clothing patterns by young consumers was obtained through an online questionnaire survey. Second, based on the IPSO and the BPNN, the nonlinear correlation mapping model between the design elements of clothing pattern and consumers’ perceptual image is established. Finally, based on the calculation of target image weight by analytic hierarchy process (AHP) method and IPSO-BPNN model, the optimal combination of clothing pattern design elements under the requirement of multi-target image is output. Taking the paper-cut pattern of sweater shirt as an example, the feasibility of this research method is verified. The research not only helped the designer to design a costume pattern that meet the individual emotional needs of consumers, but also provided a clear design index and reference, and made the costume design process more targeted, precise, and intelligent.

Keywords

• kansei engineering
• ipso-bp neural network
• paper-cut pattern
• sweater shirt
• intelligent recommendation
• personalized design